Asset Criticality Assessment
Focus maintenance where failure hurts most
Every plant has forty machines and a PM budget for fifteen. The usual way of deciding which ones get attention first is a High/Medium/Low tag someone typed in three years ago, or a gut call from whoever shouts loudest in the Monday review. Neither survives a conversation with finance when you ask for ₹18 lakh to replace a compressor.
AssetAI keeps the simple criticality tag for filtering and reporting, but the actual repair-or-replace decision runs on a separate, visible calculation called RRR — Repair, Review or Replace. It's arithmetic, not a black box, so you can put it in front of a plant head or a CFO and defend every point.
Criticality: a tag, kept honest
High, Medium, Low, or blank — that's it. There's no FMECA workshop, no severity-times-occurrence-times-detection matrix, no weighted questionnaire generating a score you then have to explain. Someone who knows the asset sets the rating by hand, the way most Indian plants actually work today.
- Criticality is a filter on the Asset and Equipment Breakdown Structure tree, so you can pull up "all High-criticality assets in Line 3" in one click.
- It's a column in the asset CSV export, useful when a reliability engineer wants to cross-tab criticality against downtime in Excel.
- A criticality master lets you maintain your own reference list — your definition of "High" for a boiler doesn't have to match a competitor's definition for a conveyor motor.
We're upfront that this is a manual tag, not a derived score. If your team needs formal RPN scoring for an audit or an ISO 55000 conversation, that's a different discipline — see ISO standards for what that entails — and AssetAI doesn't pretend to replicate it.
RRR: the number that answers "repair or replace"
RRR looks at five things the system already has data on, because you've already logged work orders, breakdowns, and downtime against the asset:
- Cumulative work-order cost against the asset's purchase cost
- Breakdown and corrective failures in the last 12 months versus the 12 months before that
- Downtime hours multiplied by the asset's downtime cost per hour
- Years since installation
- Whether the asset is still under warranty or AMC
Each factor adds or subtracts points on a fixed scale. Maintenance cost at 50% or more of purchase price adds 3 points, 30% or more adds 2. Failures that are accelerating — this year's count higher than last year's, and at least 3 — adds 2; if they're not accelerating but you've still had 6 or more in 12 months, that adds 1. Downtime losses at 25% or more of purchase cost add 2. Ten or more years in service adds 1. Live warranty or AMC coverage subtracts 1.
Add it up: 5 or more is Replace, 3 or more is Review, anything less is Repair. No hidden weighting, no AI model deciding the plant's capex — just the six inputs above, shown as plain-English lines on the card: "Maintenance cost is 54% of purchase cost (+3)", "3 breakdowns this year vs 1 last year (+2)", and so on. When someone from finance asks why a 12-year-old blower is flagged for replacement, you read the lines out loud.
Where the card lives
The RRR card sits on the asset's History screen, next to MTBF, MTTR, availability, and year-to-date downtime — the numbers a maintenance head already checks before a monthly review. It's computed live when you open the screen, not written back onto the asset record automatically. That's deliberate: an RRR verdict from six months ago, silently sitting in a field, is worse than no verdict at all. You look, you get today's number, you act or you don't.
The one number that makes or breaks it
If purchase cost is blank, the cost-ratio part of RRR is skipped, and the card tells you so plainly rather than guessing. The verdict still runs on the remaining factors, but it's weaker — a machine could be racking up huge repair bills relative to what it cost, and you'd never see that 3-point swing. For plants where old assets were entered into the asset registry without a purchase value, filling this one field in is the single highest-leverage thing you can do before trusting the RRR card for a capex decision.
Repeat offenders, flagged before the third breakdown becomes routine
A daily sweep watches for assets that keep coming back — the default is 3 breakdowns within 30 days, configurable per company if your process is stricter or looser. Alerts are de-duplicated to once per subject per week, so a chronically troublesome pump doesn't bury the maintenance head's inbox. The alert doesn't just say "this machine failed again" — it points straight at the RRR card, so the person reading it can see immediately whether the pattern has crossed into Review or Replace territory, rather than filing it under "usual trouble with the old compressor."
Who this is built for
This is for the maintenance head who has to walk into a budget meeting with a number finance will accept, and for plants trying to decide where a limited PM budget goes first — the ₹5 lakh available this quarter, and eleven assets that could each use it. If your team runs formal FMECA or RPN scoring as part of a reliability program, RRR isn't a substitute and doesn't try to be; the industries and use-cases pages go into where each approach fits. What RRR gives you instead is a number every stakeholder — maintenance, finance, plant head — can trace back to the same five facts about the same asset, in a country where spares lead times, contract labour turnover, and load-shedding already make enough decisions unpredictable without adding an opaque score to the pile. For background on how this fits into a broader maintenance system, what is a CMMS covers the basics, and concepts like OEE or TPM sit naturally alongside it. If you'd rather see the RRR card on your own asset data than read about it, book a demo and bring your worst-performing machine.
Getting the tagging and the arithmetic right on paper is one thing; running it across a few thousand assets on a real shop floor is another. The sections below cover how to set this up sensibly and where plants typically get it wrong.
Setting Up Criticality on an Indian Plant Floor
There's no prescribed matrix, which is deliberate — plants across industries from auto components to pharma to cement have wildly different definitions of what "critical" means. A single-line bottling plant might mark every filler High; a multi-line assembly plant might reserve High for anything that stops a whole shift. AssetAI doesn't force a scheme on you, but a few practices keep the tag useful rather than decorative:
- Maintain the reasoning behind each rating outside the tool — a one-line note on why an asset is High, kept in the criticality master, saves an argument later.
- Revisit ratings when a process changes, not on a fixed calendar. A machine that was Medium before a new product line often becomes High overnight.
- Don't leave the field blank by default. Blank is a valid state, but it also means the asset drops out of every High-criticality filter and report until someone sets it — easy to forget on assets added mid-year.
- Keep the criticality master short. A long list of custom categories defeats the purpose of a quick filter on the EBS tree.
None of this requires a workshop or a consultant-led exercise — it's a judgment call made once and revisited occasionally, which is exactly why AssetAI keeps it separate from anything computed.
Reading an RRR Verdict Before You Act on It
RRR gives a number and a band, but the reasoning lines are the part worth reading, not the label. A "Replace" verdict driven mostly by accelerating failures reads differently from one driven by warranty status and age — same score, different conversation with finance. Before treating a verdict as final:
- Check whether purchase cost is filled in. If it's blank, the cost-ratio component is skipped entirely and the card says so — a verdict computed without it is only using part of the picture.
- Look at whether the failure count is accelerating (recent 12 months worse than the prior 12) versus just chronically high. The first points at root-cause work; the second at a maintenance strategy problem.
- Notice if warranty or AMC coverage is quietly pulling the score down. A subsidized repair now doesn't mean the underlying trend has improved.
This is also why the repeat-offender sweep exists — it's not another algorithm, just a daily check for assets crossing a failure-count threshold in a short window, pointing whoever gets the alert straight at the RRR card instead of a new dashboard to interpret.
Where This Fits in a Broader Reliability Program
RRR and criticality tagging aren't a substitute for a maintenance strategy — they're inputs to one. Plants running TPM or tracking OEE will find RRR verdicts line up naturally with loss categories already being tracked, since downtime cost and failure frequency feed both. If your plant is working toward ISO-aligned asset management practices, the ISO standards library is a reasonable reference point, and our own standards page covers how AssetAI's records map to common audit requirements. For teams evaluating whether this level of rigor fits their plant, booking a demo is the fastest way to see the RRR card against your own asset data.
Asset Criticality Assessment FAQs
How does AssetAI actually decide whether to repair or replace an ageing asset?
AssetAI runs a separate calculation called RRR — Repair, Review or Replace — that operates independently of the criticality tag. Instead of folding a High/Medium/Low label into a hidden score, RRR takes the maintenance and cost inputs you already track and lays out the arithmetic step by step, so every figure feeding the repair-or-replace call is visible and can be checked line by line. That means when a plant head or CFO questions why a compressor is flagged for replacement, you're walking through simple addition and comparison, not defending a black box. The criticality tag still helps you locate which assets to examine first, through the Asset and Equipment Breakdown Structure tree, but the actual capex decision runs on RRR, kept deliberately apart so the two never get confused.
What's the difference between an asset's criticality rating and its RRR score?
Criticality is a manual label — High, Medium, Low or blank — set by hand and used only for filtering and reporting, while RRR is the actual repair-or-replace arithmetic used for capital decisions. AssetAI keeps the two apart on purpose: criticality tells you where to look first, not what to spend money on, and blending them would turn a simple sorting tag into an unverifiable score. Someone who knows the asset assigns criticality the way most plants already do, without an FMECA workshop or occurrence-times-detection matrix. RRR, by contrast, runs on its own transparent logic, so you can show exactly why one asset is flagged for repair and another for replacement. This separation is explained further in the CMMS basics for anyone new to the platform.
Do I need to run an FMECA workshop before setting up asset criticality in a CMMS?
No — criticality assessment in AssetAI is a manual tag, not a derived score, so there's no severity-times-occurrence-times-detection matrix or weighted questionnaire required first. Someone who already knows the asset — an engineer, a shift supervisor, a reliability lead — simply marks it High, Medium, Low, or leaves it blank, mirroring how most Indian plants already assign criticality rather than through a formal reliability study. If your plant follows structured ISO reliability practices elsewhere, you can still record that outcome as a criticality tag in AssetAI — the platform doesn't require you to build or maintain the underlying workshop scoring just to use the tag for filtering and reporting.
Can I customize what "High criticality" means for different asset types, like a boiler versus a conveyor motor?
Yes — AssetAI includes a criticality master that lets you maintain your own reference definitions instead of one fixed scale. Because a boiler failure and a conveyor motor failure don't carry the same risk or cost, the master lets each plant or asset class define its own meaning for High, Medium, or Low, rather than forcing every asset through a generic definition. This keeps the tag useful for filtering — for example, pulling up all High-criticality assets in a line through the Asset and Equipment Breakdown Structure tree — without pretending one universal standard fits every equipment type. Definitions live in the master list, so if your team's understanding of "High" changes over time, you update the reference rather than re-tagging every asset from scratch.
How can I use AssetAI's criticality data to justify a replacement budget to finance?
You use it as a filter to shortlist the relevant assets, then let the separate RRR calculation carry the actual financial argument. Criticality alone — a High/Medium/Low tag — isn't built to justify spend; it just narrows down which assets deserve a closer look, including through a column in the asset CSV export that a reliability engineer can cross-tab against downtime in Excel. The figure finance actually reacts to comes from RRR, which lays out the repair-or-replace arithmetic in the open so every number can be questioned and defended. In practice, that means pulling the High-criticality list, running RRR on those assets, and presenting that transparent calculation instead of a three-year-old tag, as covered under features.
Can criticality ratings be exported or used in reports outside the CMMS dashboard?
Yes — criticality appears as a column in the asset CSV export, so it leaves the platform in a plain, usable format rather than staying locked inside a dashboard. A reliability engineer can pull that export into Excel and cross-tab criticality against downtime, spares consumption, or any other data already being tracked, without needing a dedicated reporting module for every analysis. This works because criticality is a manual tag rather than a weighted score, so there's nothing complex to preserve on export — just the label your team assigned, attached to its asset. It sits alongside other reporting options described under features, useful for teams checking criticality against equipment performance trends over time.
If I mark an asset as "High criticality" but it has zero downtime history, will AssetAI still flag it for preventive maintenance?
Yes. Criticality is independent of failure history. A asset may have high criticality due to safety risk or production bottleneck impact, not past failures. AssetAI will recommend PM based on criticality rating regardless of downtime records. See Risk-Based Maintenance Planning for how criticality drives intervention thresholds.
See Asset Criticality Assessment on your own machines
A 30-minute demo on your plant, not our slides.